Multivariate analysis of variance

From Wikiversity
Jump to: navigation, search
Ryanscontribs.svg Resource type: this resource consists of notes.
  1. The best choice is a set of DV’s correlated with one another, but not overly so.
  2. When there is little correlation among DV’s univariate F is acceptable.
  3. Unequal cell sizes and missing data are problematical for MANOVA.
  4. Low power can mean a non-significant multivariate effect but one or more significant Univariate F’s!

Repeated-measures MANOVA[edit]

Extension of MANOVA, e.g., involving:

  1. Multiple DVs (e.g., Social, Campus, and Teaching/Education Satisfaction)
  2. Multiple Occasions (e.g., 1st, 2nd, 3rd year)
  3. Any number of IVs (e.g., Gender)
  4. Of interest are:
    1. Main effects of the multiple DVs
    2. Main effect of the multiple occasions
    3. Main effects of IVs
    4. Interactions b/w IVs and the DVs.

See also[edit]

Wikipedia-logo.png Search for Multivariate analysis of variance on Wikipedia.